IDDM registries are now being established in more than 20 areas of China, according to the standardized methods proposed by the WHO DiaMond Project, to assess the intra-country variation in disease incidence. Rates appear from 0.1/100,000 per year in Sichuan to 2.0/100,000 per year in neighboring Hong Kong. Such a large difference in risk may reflect the tremendous ethnic, cultural and genetic diversity within China. Alternatively, it may result from incorrectly classifying children with insulin-using NIDDM as IDDM. The establishment of population-based registries throughout a low risk heterogenous country, such as China, affords a unique opportunity to extend the aims of a currently funded NIH grant (RO1 DK42316) which is the foundation of the DiaMond Molecular Epidemiology Sub-Project. The proposed research will evaluate the clinical characteristics and potential genetic (HLA class II alleles), environmental (infant nutrition) and immunologic (antibodies to glutamic acid decarboxylase (GAD) and thyroid autoantigens) determinants of childhood diabetes in 15 Chinese populations by addressing the following Specific Aims: A. To examine the clinical characteristics at onset and the early natural history of diabetes in children identified from the standardized IDDM incidence registries being established for the WHO DiaMond Project in China for the years 1989-1993. B. To determine the degree to which specific HLA class II alleles and genotypes are associated with IDDM and insulin-treated NIDDM in children from selected Chinese populations using the standardized methods developed for the DiaMond Molecular Epidemiology Sub-Project. C. To assess the prevalence of auto-antibodies to GAD, thyroid peroxidase and thyroglobulin among Chinese insulin-treated NIDDM patients, IDDM cases and unrelated non-diabetic controls, and to compare these findings with those obtained for Caucasian IDDM cases and non-diabetic controls from Allegheny County, PA. D. To evaluate the interactions between HLA class II alleles and auto- antibodies to GAD, thyroid peroxidase and/or thyroglobulin among Chinese IDDM and insulin-treated NIDDM patients and to compare these findings with those obtained for Caucasian IDDM cases.